Markets Weigh Impact of AI on Software Sector

What happened?

• Volatility among global technology stocks, particularly software stocks, has increased over recent days. These stocks have been hit by investor concerns that advances in agentic AI call into question the viability of traditional software company models.
• Credit markets have also reacted to concerns about levels of leverage and implications for software-related loans.
• Investors across asset classes are evaluating the impact on the software sector of several dynamics, including:
  • New forms of artificial intelligence have the potential to redefine almost every industry. The introduction of Anthropic’s Claude Code, for example, may be viewed as an inflection point, illustrating that traditional software development economics are no longer a given.
  • AI creates the potential for companies to write their own software and replace third-party tools.  Generative AI tools make coding simpler, which could lower switching costs for enterprise software customers and increase the risk of client churn.
  • These tools could lower the cost of software development and change the competitive environment.
  • The near-zero marginal cost of work tasks executed by AI could reduce the perceived value of the software tools that helped to enable this output in the past.
  • If an AI agent can execute a multitude of tasks across different tools, enterprise customers may need fewer software solutions and seat licenses, empowering them to push back on pricing.

What are the market implications?

• Software and SaaS (Software as a Service) stocks have experienced heavy falls in recent days, and the S&P 500 Software and Services sector index is over 15% lower year to date.¹
• In fixed income, the weakness in technology loans, where software and IT services is greater than 15% of the index², has extended to bonds of software companies in high yield.

What are the developments to watch?

Equity

Dom Rizzo, Portfolio Manager, Global Technology

• We are closely monitoring AI’s progress in advancing from assistive solutions that help users accomplish tasks to powerful tools that increasingly can execute more complex workflows. The key concern is no longer whether agentic AI boosts productivity, but whether it reshapes traditional application vendors’ business models and economics.
• Software isn’t dead. However, uncertainty is high. And where long-term value accrues in the software industry may be shifting. Traditional software companies’ long-term pricing power has increasingly come into question. Applications that are generic, or where the main value is helping users with repeatable tasks, could be more at risk. Controlling unique data, platforms, or distribution could be important differentiators. We are also monitoring whether some application vendors will be able to evolve their business models to monetize the increased usage of software tools by AI agents.
• As software stocks look for a bottom from here, the market is likely to anchor on enterprise value to free cash flow (adjusted for stock-based compensation) and whether a company meets the “rule of 40”—where its free cash flow margin and revenue growth should add up to more than 40% to indicate strong financial health. The challenge is that software valuations—including those of companies viewed as best in class—don’t appear that compelling yet, especially when you consider the long-term uncertainty about where value accrues. At this juncture, the risk/reward setup in the hardware and equipment companies exposed to the massive spending on AI infrastructure seems much cleaner than in application software. Valuations for hardware and equipment companies also still appear reasonable.
• In this uncertain environment, it is important to stay nimble and maintain a strong framework. We are looking for software and other companies with linchpin technologies that are innovating in secular growth markets, boast improving fundamentals, and trade at reasonable valuations. The current selloff underscores why it is critical to know the kinds of businesses you own. Markets are treating the sector indiscriminately by failing to distinguish between companies that are structurally vulnerable to disruption and those with the scale, balance sheets, and strategic optionality to adapt to a new environment.

Fixed Income

Michael Trivino, Associate Portfolio Manager, High Yield

• In fixed income, the narrative that AI may displace traditional software remains a dominant theme, particularly in the loan market, where technology loans have sharply underperformed year to date. While concerns about certain vulnerable credits are understandable, the weakness has extended to companies we view as higher quality and potentially more resilient to AI disruption.    
• It’s important to recognize that software is deeply embedded, costly to replace, and requires maintenance and updates. It is not simply about code. The timeline of displacement is likely longer than markets anticipate given multiyear contracts, compliance, and training that will need to take place. While sentiment and enterprise valuations in the sector have deteriorated, fundamental earnings and cash flow are unlikely to materially change much in the near term. 
• We are monitoring the sector very closely and have developed a framework to evaluate AI displacement risk and categorize securities to help assess potential downside and take advantage of opportunities during this volatile period.

Private Equity

David DiPietro, Head of Private Equity

• In the private markets, companies which appear well-positioned are those that provide the core technology powering artificial intelligence. This includes businesses that build the data and computing infrastructure behind AI, as well as companies developing the most advanced AI models. These businesses tend to benefit broadly as AI adoption continues to grow, regardless of which specific products or use cases ultimately succeed.
• Selectivity is vital when evaluating AI-enabled software applications. Considering an underwriting horizon³ of three to five years, it is harder to assess how durable these businesses will be over time.  As the underlying AI technology continues to improve, features that once differentiated standalone applications may become widely available or built directly into larger platforms, potentially reducing their long-term value.

Private Credit

Alan Schrager, Senior Partner and Portfolio Manager, Oak Hill Advisors

• Over the past five to 10 years, technology and software companies have grown to represent approximately 15% to 20% of activity in the private credit and syndicated bank loan markets⁴. The sector’s recurring revenues and predictable cash flows have made it especially attractive for private equity sponsors pursuing leveraged buyouts (LBOs), with financing provided through bank loan/collateralized loan obligation (CLO) origination and private credit channels. While some risk of disruption exists, a number of better-positioned software companies are deeply entrenched in their customers’ operations, with high switching costs that make displacement unlikely within the typical loan maturity period.
• In today’s environment, investors are becoming much more discerning regarding business, product, and credit quality. This underscores the importance of a fundamental, bottom-up research approach. It is increasingly important to separate signal from noise and to distinguish winners from losers. In the near term, we are closely monitoring churn levels and margin pressures but continue to see opportunities where valuations appear oversold.